How do I diagnose a metallographic prep problem?
Almost every preparation defect comes from one of three root causes: insufficient damage removal from an earlier step, mechanical disturbance at the surface that has not yet been polished or etched away, or contamination introduced between steps. Before you change anything, identify which of those three you are looking at.
A useful habit: after each step, rotate the sample 90° before starting the next. New scratches at the previous orientation tell you the prior grit's damage isn't fully removed; that's the single most common reason a "polishing problem" persists no matter how long you polish.
The 3× rule
As a starting point, polish at each step for at least three times the time required to establish a uniform scratch pattern at that step. Stop when the previous step's pattern is fully gone; verify by rotating between steps.
Residual Scratches from Prior Step
Long, parallel scratches all oriented in the same direction, recognisable as coarser than the abrasive currently in use. They were generated by the prior grit and the current step hasn't yet removed them.
Same-direction scratches you can't polish out
CommonSymptom: Scratches all running one way, deeper than what the current abrasive should produce. They survive long polishing times.
- Insufficient time at current step
- Polish at least 3× the time needed to establish a fresh scratch pattern at this grit. Verify by rotating the sample 90° between steps.
- Skipped a grit in the sequence
- Keep the standard SiC sequence (e.g. 240 → 320 → 400 → 600) and don't skip more than one step. For diamond polishing, don't jump more than a 3× ratio in particle size.
- Pad worn, glazed, or loaded
- Inspect the pad; a glossy, hard surface no longer cuts. Dress with a conditioning stick or replace the pad.
- Force too low
- Typical metallographic force is ~5–6 lbf (~25 N) per 1.25″ sample. Too-light force just polishes the disturbed layer instead of cutting fresh material.
- Final step skipped 90° rotation
- Without rotating between steps you cannot verify scratch removal. Make it part of the procedure.
Deep gouges or isolated heavy scratches
WorkflowSymptom: One or two scratches significantly deeper than the rest, often longer than the field of view.
- A loose particle of coarse abrasive on the pad
- Return to the grit that could have caused the gouge and re-polish. Then rinse equipment thoroughly and start the finer step on a clean pad.
- Damage from sectioning never fully removed
- If a deep scratch appears on the first grinding step and won't budge, the cut left subsurface damage. Drop one grit coarser or extend grinding before progressing.
Newly Introduced Scratches
Random-direction scratches, or scratches noticeably finer/coarser than the abrasive in use, appearing on a step that was previously clean. The cause is almost always a foreign particle, either carry-over from a coarser step, dropped onto the pad, or shed by a worn pad.
Random fine scratches mid-polish
CommonSymptom: Short scratches in random orientations appearing during a fine polish step, often after switching from grinding.
- Carry-over of coarse abrasive on the sample, hands, or holder
- Ultrasonic-clean (or thoroughly rinse and brush) the sample and holder between every step. Use a separate brush for each grit.
- Coarse abrasive carried in the lubricant or by airborne dust
- Keep pads covered when not in use. Don't store coarse and fine pads in the same drawer uncovered. Filter recirculating coolant.
- Pad contaminated from prior step
- Dedicate each polishing cloth to a single grit. Never reuse a cloth at a different abrasive size.
Scratches deeper than the current abrasive should produce
MaterialSymptom: Scratches at the polishing step that look the size of an earlier grinding grit.
- Embedded SiC in soft metal (Al, Cu, Pb)
- SiC fragments can lodge in the matrix and re-scratch downstream. For soft metals, finish grinding on finer SiC at light force, or switch to diamond on a rigid disc.
- Pad shedding from worn or damaged cloth
- Replace the cloth. Inspect under glancing light before each batch.
Comet Tails
Tail-shaped streaks emerging behind hard particles, inclusions, or pores, all pointing in the same direction (the direction of relative motion between the sample and the pad). They form when loose abrasive is dragged across the surface around an obstacle.
Streaks behind hard particles, all pointing one way
CommonSymptom: Tail-like streaks behind every hard phase or inclusion at the final polish.
- Pad nap too long
- Switch to a napless pad (e.g. TEXPAN) or a low-nap pad (e.g. ATLANTIS, GOLDPAD) for the final step on samples with hard inclusions.
- Too much slurry on the pad
- The pad should be moist but not wet. Reduce dose and let the slurry work; flooding the pad floats abrasive across the surface.
- Suspension has settled
- Shake colloidal silica and alumina suspensions vigorously before every use. Settled abrasive is just water.
- Force too high
- Drop to 3–4 lbf per sample on the final step. Tails sharpen as force rises.
- Hand-held free polishing
- Use a sample holder so the relative motion stays planar; free-hand polishing varies pressure unpredictably and amplifies tails.
Smearing & Disturbed Layer
The polished surface looks bright but bland: grain boundaries are faint or absent, second phases are blurred, the structure looks "smeared." A thin layer of mechanically deformed metal (the disturbed or smeared layer, historically called the Beilby layer) is masking the true microstructure. This is the defining problem for soft metals.
No visible structure on soft metals after polish
Aluminum • Copper • Lead • Tin • GoldSymptom: Etch develops poorly or unevenly; features look washed out; phases blur into the matrix.
- Pure mechanical polish on a soft alloy
- Finish chemo-mechanically (0.04 µm colloidal silica or 0.05 µm Nanometer alumina) on a soft synthetic-nap pad. The chemical action removes the disturbed layer the mechanical step alone cannot.
- Final-stage force too high
- Use 2–3 lbf for the chemo-mechanical step on soft metals; let chemistry do most of the work.
- Disturbed layer too thick to remove in one pass
- Polish-etch-polish: brief final polish, light etch to reveal damage, repeat. Each cycle removes a thinner disturbed layer than the last.
- Coarse grinding damage carried into polishing
- Add a fine grinding step (1200 or 2400 SiC) before going to diamond, especially for pure aluminum and pure copper.
Pull-Outs & Plucking
Dark holes where particles, inclusions, or graphite should be — sometimes called particle loss, plucking, or pluck-out. The mechanical action of the pad has torn material out of the matrix. A pull-out is recognisable by a smooth-bottomed dimple, distinct from a real corrosion pit (sharp, irregular edges) or porosity (often interconnected, irregular shape).
Graphite missing or shrunken in cast iron
Cast ironSymptom: Graphite nodules appear smaller than they should, with rounded edges, or are entirely missing.
- Long-nap final pad with too much abrasive
- Finish on a short-nap or napless pad with 1 µm diamond. Stop short of any nap-velvet alumina or colloidal silica final step that would round the graphite edges.
- SiC used too late in the sequence
- For cast iron, transition to diamond on a rigid composite disc after the fine grind. SiC at fine grits can fracture and embed in graphite.
- Excessive polishing time
- Limit each polish step to 2–4 min. Long times round the graphite-matrix interface.
Inclusions pulled from steel (MnS, oxides)
Steel inclusion ratingSymptom: Holes where ASTM E45 inclusion rating would expect particles; rated severity comes out too low.
- Pad nap grabbing inclusions
- Use a short-nap or rigid composite pad for the final polish. Long-nap velvets are not suitable for accurate inclusion rating.
- Excess force on final
- 3–4 lbf for inclusion-rating preparation.
- Final-step time too long
- Keep diamond final to 2 min and any alumina/silica polish to ≤1 min when retention matters.
Particle pull-out in porous or brittle samples
Powder metal • ceramics • coatingsSymptom: Particles missing from sintered or porous materials; cermet hard particles plucked from binder.
- No support behind the particle
- Vacuum-impregnate porous samples with a low-viscosity epoxy before grinding. The cured resin supports particles from below.
- Aggressive coarse grind fractures particles before they are supported
- Start finer than usual (320 or 400 SiC) on impregnated porous samples to minimise initial damage.
Relief & Edge Rounding
Relief is differential material removal between phases of different hardness: hard particles or coatings stand proud, the softer matrix sits below. Edge rounding (also called edge dropoff, edge retention loss, or rounded edges) is the same effect at the boundary between sample and mounting compound, and it is the enemy of any analysis at the edge: case hardening, coatings, weld interfaces.
Hard phases proud of the matrix; halo around inclusions
CommonSymptom: Loss of focus at high magnification between phases; visible step under DIC; carbides "shadowed."
- Pad nap too long
- Switch the final step to a napless or short-nap pad. Pad nap is the single biggest driver of relief.
- Excessive final-step time
- Cap final polishing at 1–3 min. Diminishing returns set in fast and relief grows linearly with time.
- Mismatched abrasive hardness
- For very hard particles in a soft matrix (e.g. tool steel carbides), diamond is mandatory; alumina is harder than the matrix but softer than the carbides and grows the step.
Rounded edges at coating, case, or weld interface
Critical for edge analysisSymptom: Edge of sample drops below the focal plane; case-depth measurements read short; coating thickness can't be measured accurately.
- Sample harder than mounting resin
- Use an edge-retentive mount: conductive phenolic (e.g. condu-met type), mineral-filled epoxy, or a hardness-matched compression resin.
- Long-nap or soft final pad
- Use a firm low-nap pad (e.g. TEXPAN, ATLANTIS) for the final polish to maintain edge support.
- Coating not protected during grinding
- For thin coatings or oxide scales, electroless-nickel-plate the surface before mounting, or mount with a hard mineral-filled epoxy that contacts the coating directly.
- Sample loose in mount
- A gap between resin and sample is fatal for edges. See the mounting section for void/shrinkage fixes.
Real Pits vs. Polishing Artifacts
Dark spots on a polished surface can be many things, and the correct response depends entirely on which. Re-polish briefly and look at the same area again: real defects stay put because they're part of the sample; artifacts move, change, or vanish.
Likely a real defect
- Sharp, irregular edges
- Reappears in the same place after re-polish
- Found in clusters consistent with corrosion or porosity
- Confirmed by etching: pit is bordered by attack patterns
Likely a polishing artifact
- Smooth, round dimple with clean bottom under DIC
- Has a streak emerging from it (it's a comet-tail tip)
- Shifts, shrinks, or disappears after re-polish
- Distributed in a direction (drag pattern)
Pitting that appears after etching
Etch issueSymptom: Surface was clean after polish; pits appear during or after etching.
- Etchant pooling and drying on the surface
- Rinse with running water immediately at the endpoint, then displace with ethanol and dry with warm forced air. Never let etchant air-dry.
- Stale etchant
- Mix fresh for critical samples; many common etchants degrade in hours.
- Galvanic effect from clip or mount
- If a metallic clip contacts the sample during immersion etching, dissimilar metals form a cell. Insulate the clip or change to swab etching.
- Etchant too aggressive for this alloy
- Switch to a milder formulation; see the etching procedures guide for alloy-specific options.
Etching Problems
Most etching defects trace back to one of three causes: a still-disturbed surface that the etchant can't read, the wrong etchant chemistry or concentration, or poor rinsing and drying technique. Diagnose by what the surface looks like, not by how long the etch was.
Over-etched: background dark, features dissolved
CommonSymptom: Grains over-developed, boundaries widened into trenches, fine features destroyed.
- Etch time too long
- Repolish through the final step and re-etch with shorter time. Most etchants reach a useful endpoint in seconds, not minutes.
- Etchant too concentrated
- Dilute fresh etchant per the published procedure rather than guessing. A 2% Nital is not interchangeable with 5%.
- Sample warm from polishing
- Let the sample reach room temperature before etching; warm surfaces etch faster than expected.
Under-etched: no contrast
CommonSymptom: Long etch produces little or no visible structure.
- Disturbed layer still present
- The most-missed cause. Repolish (extend final polishing or add a chemo-mechanical step), then re-etch.
- Etchant exhausted or expired
- Mix fresh. Color and clarity drift; potency drops well before either is obvious.
- Wrong etchant for the alloy
- Check the published etchant for your alloy class. Nital is for plain-carbon and low-alloy steels; austenitic stainless needs Marble's, Glyceregia, or Vilella's; aluminum needs Keller's or Tucker's.
- Surface contaminated with oil or fingerprints
- Rinse with ethanol immediately before etching; never handle the polished face.
Stains, water marks, brown tinting after etch
Drying techniqueSymptom: Yellow, brown, or rainbow patches; outlines of water droplets; uneven darkening.
- Etchant or rinse water allowed to dry on the surface
- Use a 3-step dry: water rinse → ethanol displacement → warm forced-air dry. Don't air-dry from wet.
- Contaminated drying air (compressed shop air with oil)
- Use a dedicated clean blower or filtered air. Oily air dries to fingerprint-shaped haze.
- Hot etchant or sample
- Cool both before drying; warm samples wick rinse fluids into the etched layer.
Uneven or streaked etching
ApplicationSymptom: One area properly etched, another under-etched; wipe marks across the surface.
- Etchant applied unevenly
- For swab etching, use enough etchant on a clean cotton swab to flood the surface; move continuously. For immersion, agitate the sample gently.
- Surface partially contaminated
- An untouched fingerprint or oil patch will resist etching. Clean with ethanol immediately before the etch.
- Residual disturbed layer in patches
- Heavy-pressure spots during the last polish leave more disturbance and etch less. Use uniform pressure or revisit the polishing technique.
Mounting Defects
Mounting defects propagate. A void, gap, or crack near the surface of interest will fill with etchant or polishing slurry, hide a real defect, or worse, round the edge of the very feature you wanted to analyse. Diagnose at the mounting stage; don't wait until polishing.
Voids around the sample
CommonSymptom: Bubbles or open cavities visible at the resin-sample interface or within the cured mount.
- Air trapped against irregular sample features
- For castable resins, vacuum-impregnate. For compression mounts, ensure full pressure is maintained throughout the cure (not just the heat-up).
- Porous sample drawing air
- Vacuum-impregnate before final mounting, or use a two-stage approach: low-viscosity epoxy fill under vacuum, then encapsulate.
- Sample surface oily or wet
- Clean and dry every sample before mounting. Castable resins won't wet contaminated surfaces.
- Compression pressure released too early
- Hold pressure through cooling on phenolic and acrylic mounts; venting early traps shrinkage voids.
Shrinkage gap around the sample
Edge analysis killerSymptom: Thin gap visible between sample and mount; etchant or polishing slurry pools at the edge.
- Excessive accelerator in castable epoxy
- Use the manufacturer's recommended mix ratio. Over-accelerated resin cures hot and shrinks more.
- Hardness mismatch between resin and sample
- For hard samples, use mineral-filled epoxy or a hard phenolic; for soft samples, a less rigid resin will move with the sample during preparation.
- Cold sample at casting
- Bring samples to room temperature before casting; a cold sample causes condensation and accelerates shrinkage on cure.
Mount cracking
CommonSymptom: Radial or circumferential cracks in the resin, sometimes through the sample.
- Cooling too fast on compression mounts
- Use the press's controlled cool cycle, or let mounts cool naturally. Fast cooling on phenolic or DAP induces thermal stress.
- CTE mismatch between sample and resin
- For temperature-sensitive samples, use castable epoxy (room-temperature cure) rather than compression mounting.
- Too much accelerator → exothermic cure
- Smaller batches, manufacturer ratio, and pre-cooling the mold cup reduce exotherm cracking on castable epoxies.
Heat & Burning
Heat damage is uniquely destructive because it changes the sample itself. A burned sample can show false temper colors, retempered martensite, recrystallised grain structure, or melted re-cast layers, all features that didn't exist before preparation. Once burned, the affected layer cannot be polished away to recover the original structure; you must remove enough material to get below the heat-affected zone or section a fresh piece.
Blue/straw temper colors after grinding
Invalidates hardness dataSymptom: Tint visible on the polished surface; hardness readings inconsistent with specification on hardened steel.
- Coolant flow inadequate at the cut/grind interface
- For sectioning, flood-cool with proper recirculation; for grinding, ensure constant water flow at the abrasive-sample contact. Switch to a finer SiC if coarse cuts overheat.
- Excess force on grinder
- Lower force; let the abrasive cut. A grinder pushed too hard generates more heat than material removal.
- Glazed grinding paper
- Glazed paper rubs instead of cutting. Replace as soon as material removal slows.
- Wrong cutoff wheel for the alloy
- Hardened steels need a softer-bonded abrasive wheel; soft alloys need a harder bond. Mismatch causes the wheel to either glaze (heat) or wear away (no cut).
Resin softening or melting during preparation
WorkflowSymptom: Smell of burning plastic; mount edge rounded faster than expected; resin smearing onto the pad.
- Insufficient coolant
- Wet grinding only; never dry-grind a thermoset or thermoplastic mount.
- Acrylic mount on a hard alloy at high speed
- Acrylic softens around 90 °C. Use phenolic, epoxy, or DAP for samples that require aggressive grinding.
- Pad clogged with resin swarf
- Clean the pad frequently or replace; a loaded pad rubs rather than cuts, generating more heat.
Pad, Cloth & Abrasive Issues
Many "polishing problems" are actually pad or cloth problems. Inspect the pad under glancing light before each batch; glazing, loading, and uneven wear are visible at a glance and explain a lot of repeat scratches, inconsistent finish, and slow cutting.
Pad glazed: surface shiny, no longer cutting
CommonSymptom: Material removal slows or stops; scratches won't progress; sample heats up.
- Embedded fine particles fill the pad surface
- Dress the pad with a diamond conditioning stick at the start of each session, or replace it.
- Insufficient abrasive replenishment
- Re-charge the pad with fresh suspension before it dries; don't try to extend life by under-feeding.
Pad loaded with metal swarf
Soft metalsSymptom: Visible smear of removed metal on the pad surface; finish quality degrades over time.
- Soft metal smearing into the pad
- Reduce force, switch to a different pad style with better swarf clearance, or shorten the step.
- Pad past service life
- Track pad usage; replace per the manufacturer's recommendation rather than waiting for failure.
Wrong abrasive choice for the material
MaterialSymptom: Embedded abrasive showing as bright specks; inability to reach a scratch-free finish.
- SiC on pure aluminum, copper, or other soft metals
- SiC fragments embed in soft matrices. Use diamond on rigid discs for grinding soft metals; finish with alumina or colloidal silica.
- Alumina on tool steel carbides
- Alumina is too soft to cut tungsten or vanadium carbides; relief grows fast. Use diamond throughout to final, and an oxide finish only briefly.
- Coarse diamond on a soft matrix
- Diamond particles can transfer and embed if the matrix is much softer than the diamond bond. Choose appropriate suspension viscosity and pad type for the alloy.
Cross-Contamination
Cross-contamination is the catch-all cause for "scratches that shouldn't be there." Most of it is preventable with simple discipline: dedicated pads per grit, separate brushes per step, ultrasonic cleaning between steps, covered storage of unused consumables.
Carry-over from coarse to fine
CommonSymptom: Random scratches deeper than the current abrasive should produce, appearing after a grit transition.
- Coarse abrasive still on the sample or holder
- Ultrasonic-clean for 30–60 s between every step; rinse the holder separately.
- Coarse abrasive still on the operator's hands or gloves
- Wash hands or change gloves between coarse and fine steps. Don't handle the polished face of the sample.
- Coarse pads stored uncovered near fine pads
- Store each pad in a covered container; never stack uncovered.
Cross-contamination between dissimilar alloys
Mixed batchesSymptom: Iron contamination on aluminum samples; spurious EDS peaks; staining patterns inconsistent with the alloy.
- Same pad used for ferrous and non-ferrous samples
- Dedicate pad sets to material families. Ferrous swarf on a copper-polishing pad will produce iron contamination and may galvanic-etch the copper.
- Same etching tongs/clips across alloys
- Use plastic tweezers or dedicated tools per alloy class for etching.
Best-Practice Checklist
Most preparation problems are prevented by a small number of habits applied consistently. This is the short list that resolves the majority of issues described above.
- Rotate 90° between steps. The only reliable way to verify the prior step's damage is gone.
- Apply the 3× rule. Polish each step for at least three times the time needed to establish a fresh scratch pattern.
- Ultrasonic-clean between steps. Single most effective contamination control.
- Dedicate pads per grit and per material family. Don't try to share.
- Shake suspensions vigorously before every use. Settled abrasive is just water.
- Inspect the pad before each batch. Glazing, loading, and wear are visible at a glance.
- Never let etchant or rinse water air-dry. Use water → ethanol → warm forced air.
- Match abrasive to material. Diamond for hard particles, oxide for soft matrices, never SiC on pure Al/Cu past fine grinding.
- Match resin to sample hardness. Edge-retentive mounts for hard samples and edge analysis.
- Document what changed. When a problem appears, log every variable; reproducibility is the foundation of troubleshooting.
Find a Procedure for Your Material
Preparation problems are easier to avoid than to fix. Start with a validated procedure for the alloy class you're working with; our material-specific procedure guides give grit sequences, force, time, pad, and etchant recommendations tested on real samples.
Frequently Asked Questions
How do I tell whether a dark spot is a real pit or a polishing artifact?
Re-polish briefly and look again at the exact same location. Real defects stay put; they're part of the sample. Polishing artifacts (pulled inclusions, comet-tail tips, residual scratches viewed end-on) shift, change shape, or disappear after another preparation cycle. Sharp irregular edges suggest a real pit or pore; smooth round dimples with a clean bottom usually mean a pulled-out particle; any streak emerging from the spot is a comet tail, not a pit.
Why do scratches from the previous step keep showing up after polishing?
The most common cause is insufficient time at the current step; spend at least 3× the time needed to establish a uniform scratch pattern at the new grit. Other common causes are skipping more than one grit in the sequence, not rotating 90° between steps (so you can't actually verify scratch removal), a worn or glazed pad that is no longer cutting, and insufficient force (typical metallographic force is ~5–6 lbf per sample). A worn cloth that looks fine will still fail to cut; inspect under glancing light.
What causes comet tails behind hard particles?
Loose abrasive being dragged across the surface around a hard obstacle. The fix has four parts: switch to a short-nap or napless cloth for the final step, reduce slurry dose so the pad is moist but not wet, shake the suspension before use so abrasive is in suspension and not settled, and lower the force to 3–4 lbf. A sample holder rather than free-hand polishing also helps because it keeps relative motion planar.
How do I prevent pull-outs in cast iron graphite or in steels with sulfide inclusions?
For cast iron, never finish on a napped cloth with SiC. Finish with 1 µm diamond on a short-nap or napless cloth, and stop short of any final velvet alumina polish that would round the graphite. For inclusion retention in steels (ASTM E45 rating work), keep the final polishing time to 2 min or less, reduce force to 3–4 lbf, and use a rigid composite or short-nap pad. Long-nap velvets actively grab inclusions and tear them out.
Why does my polished aluminum or copper sample look smeared with no microstructure?
Soft metals develop a mechanically disturbed surface layer during polishing that hides the true microstructure. Pure mechanical polishing can't remove that layer reliably. Finish chemo-mechanically: 0.04 µm colloidal silica or 0.05 µm alumina on a soft nap pad, very light pressure (2–3 lbf), then etch. For stubborn cases, polish-etch-polish in cycles; each cycle removes a thinner disturbed layer than the last.
What causes voids in compression and castable mounts?
In compression mounts: insufficient pressure during cure, releasing pressure too early, or surface contamination that the resin can't wet. In castable epoxy mounts: trapped air around irregular features, or porous samples drawing air. Use vacuum impregnation for porous samples (foams, sintered parts, oxide scales, ceramics), warm the sample slightly before casting, choose a low-viscosity epoxy that flows into fine features, and hold compression mounts under pressure all the way through cool-down, not just heat-up.
Why is my sample showing brown stains or water marks after etching?
Stains form when etchant or rinse water dries on the surface. Use a three-step dry sequence every time: running water rinse to flush all etchant, ethanol displacement to push water off, then warm forced air. Never let the surface air-dry from wet. Don't use compressed shop air without a coalescing filter (oil) and don't use a contaminated towel. If stains have formed, repolish lightly through the final step and re-etch.
Why is the edge of my mounted sample rounded under the microscope?
Edge rounding comes from a hardness mismatch between sample and mount, or from a soft long-nap pad on the final polish. Use an edge-retentive mounting compound such as conductive phenolic, mineral-filled epoxy, or a hardness-matched compression resin. Polish on rigid composite or short-nap pads, keep final polishing time short (1–2 min), and for the most demanding edge work (thin coatings, oxide scales, case depth measurement) consider electroless nickel plating the surface before mounting.